Question Shearwater, Subsurface & Gas Switches - Advice?

[Badger7]

Hey guys,
I'm usually pretty good with figuring out software configurations, but this one is vexing me. The wrong cylinder is being shown on each side of the gas switch, and all attempts to fix it have failed miserably. This isn't just a single dive, any dives with a gas switch are having the same problem in Subsurface.

When diving a shearwater perdix2 in OC Tech mode with two gases defined and doing a gas switch on ascent at the appropriate depth, everything goes fine with the dive, deco is calculated as expected, and technically everything works as expected. I just can't get the dive to look right after-the-fact when imported into subsurface. Single gas dives work just fine, even in the same OC Tech mode, but the gasses on the graph seem to be getting flip-flopped when a gas switch is used during the dive.

For reference, I have a AI transmitter on back gas (the EAN26 in this case) and no transmitter on the deco cylinder (EAN50). Start and end pressure for the EAN50 were just read off the spg and entered manually into the equipment tab in subsurface. I see that you can right click on the gas switch on the graph itself to identify the other cylinder in the gas switch, but instead of actually flip-flopping the gas for dive, it intead treats it as if both cylinders are used simultaneously. No matter what I've tried under the equipment tab, I cant get the EAN26 back gas to be respresented as being used on the working portion of the dive and the EAN50 to be shown as being used on ascent during decompression. Behaves the same on both Android and Windows, both synced to cloud and showing the same dive log.

I am 1000% positive I had the perdix set with EAN26 on descent and set the gas to EAN50 on ascent at 70ft, but Subsurface displays them flip-flopped.

These are the two options when toggling the "edit gas change" by right clicking on the gaschange event on the chart:

looks like this:

looks like this:

This is how the gases are set up in the equipment tab of that dive in Subsurface:

Any ideas on what I'm missing? I really don't want my dive logs to look like I'm taking EAN50 to depth, hopefully I'm just missing a small detail, or is this some type of known bug/nuance of subsurface?

 

[jborg]

Odd. Are you on a modern version of Subsurface?

 

[Mobulai]

Any ideas on what I'm missing? I really don't want my dive logs to look like I'm taking EAN50 to depth, hopefully I'm just missing a small detail, or is this some type of known bug/nuance of subsurface?

Seems you might uncovered a real bug

I had the same exact situation with SPg2, transmitters on bottom gas (sidemount not BM) and spg on 50% and that works quite alright out of the box (opposite to your case)

On CC however — when I prompted set point changes on the G2 (I use it as the backup, so I have a low and high set points to account for various situations like flushes) the logs go bananas (basically reads as OC gas changes from air to o2)

I can’t tell what is the bug (due to lack of reproducibility and ability to compare — I don’t have the Petrel logs as it is a rental unit); but I assume it has to do with the data structure of the logs

We might be referring to similar bugs, or not
Eitherway I’ll keep an eye here on the thread

 

[Badger7]

Odd. Are you on a modern version of Subsurface?

Good question, on windows im using v6.0.5214, but there is a 6.0.5404 update available so I'll update and review

 

[jborg]

Sounds new enough I guess, just I don't see this on my Subsurface. Behaviour seems to be that the first gas in the list is used up to the first gas switch point (as you'd expect)...

 

[Mobulai]

Sounds new enough I guess, just I don't see this on my Subsurface. Behaviour seems to be that the first gas in the list is used up to the first gas switch point (as you'd expect)...

So.. if I have the 50% 1st on the list (on the dive computer) it just assumes that this is the “bottom gas” kinda thing?
I’d assume the assignment of the transmitter overrides that (might be a hint at what went wrong)

 

[Badger7]

So two things
1) Minor differences in the latest version, adds a new information window that follows the cursor as it moves across the graph, thats cool, but no change in the gas switch behavior.

2) Maybe its working as intended, but not just as I expected. When I set the gas switch as in the second example, with the AL40 selected for the gas change, yes it shows the pressure of each cylinder stretched across the entire dive, but thats actually what happens in reality, isn't it? It even shows the increase in pressure on the back gas after I stopped using it, but the water temp changed froim 47 to 57 degrees on ascent, increasing the tank pressure from 1924 to 1988 PSI as the gas warmed back up. It would be nice if it showed the starting tank pressure on the deco bottle at the gas switch, but since there isn't any air integration on that cylinder, it really can't be expected to know that, but it would be a nice assumption.

Is this how it is supposed to look? The more I stare at it, maybe this is correct after all?

If so, new question, any idea if there is a way to push the graph of the deco cylinder tank pressure to start at the gas switch instead of starting at 0 minutes? I suppose I could live with it as-is, and its totally not worth having a AI transmitter hanging off a deco bottle just to make the graphs look right, not when the SPG is right next to valve hanging visible the whole time right under my left arm. The half-off sale right now on Swifts almost makes me want to do it, but I'm controlling myself as best as I can.

 

[Mobulai]

well this is what I get (with the same conditions, mixed AI & SPG usage)
the 2 redlines are my Air sidemount tanks

btw I can't view your attachment in the last post

 

[jborg]

I don't think so... I don't actually have a dive with a computer gas switch in it, but adding it manually goes from a dive with one gas:

Adding an al40 with no gas switch shows both tanks concurrently:

Adding a gas switch shows the main tanks up until that point (because there is a transmitter and the data is timestamped, I presume) and the second tank from that point onwards:

 

[gbirch]

So.. if I have the 50% 1st on the list (on the dive computer) it just assumes that this is the “bottom gas” kinda thing?

The order of gasses on the list on the Shearwater is not fixed. The computer will always reshuffle the list to put the gasses in descending order of %O2 regardless of activation status. For example, in System Setup, an inactive 80% O2 will list above an active 50% which is above a 26% ...